Artificial intelligence (AI) holds great potential for tech giants like Microsoft and Alphabet, but both companies are anticipating deeper investments before reaping significant bottom-line gains.
Microsoft has witnessed a sharp rise in costs while building new data centers to support AI. Its capital expenditures will continue to increase as it acquires chips from companies like Nvidia to power these data centers. The AI costs are twofold for Microsoft: first, to enhance its own products, including the upcoming Copilot AI assistant; second, to cater to companies utilizing its Azure cloud computing services to create AI-based products. Microsoft’s executives expect the service to generate the majority of its revenue in the latter half of fiscal year 2024, ending June 30.
Alphabet, on the other hand, has initially managed to keep costs down, but it may not be the case for long. The CFO, Ruth Porat, stated that delays in data center construction impacted second-quarter capital expenditures, leading to lower-than-expected spending.
Google’s advantage lies in having its custom chip called the Tensor Processor Unit (TPU), designed to handle AI work, which helps lower costs. However, Google also acknowledges that it will continue to purchase chips from other companies alongside using its proprietary TPUs, potentially affecting profit and growth.
While Microsoft’s aggressive acquisition of Nvidia chips (due to the absence of its own silicon alternatives) is driven by investors’ desire to witness faster progress, Google’s approach, leveraging its custom chips, has contributed to cost management. Nonetheless, both companies emphasize that the AI investment inflection point is yet to be fully realized, indicating their long-term commitment to harnessing AI’s potential.